K&C Science Report – Phase 1 Synergetic Use of ALOS PALSAR, ENVISAT ASAR and Landsat TM/ETM+ Data for Land Cover and Change Mapping

Interferometric ALOS PALSAR Fine Beam Single/Dual, multi-temporal ENVISAT ASAR Alternating Polarization, and Landsat-5 TM/-7 ETM+ data are used for the generation of land cover and change maps. In synthesis, the product generation foresees two main steps: The first one consists in a rigorous data pre-processing, including interferometric processing (PALSAR), geometric / radiometric calibration (PALSAR, ASAR, TM/ETM+), and multi-temporal speckle filtering (PALSAR, ASAR); as result, terrain geocoded coherence, sigma nought, and top-of-atmosphere reflectance products are obtained. The second part is dedicated to data classification and fusion. Classification is performed by means of a prior knowledge-based approach exploiting interferometric, multi-temporal intensity, and spectral signatures. Data fusion and change detection are subsequently applied at semantic level. Results – based on acquisitions over Malawi (country-wide) and on an area in Brazil – show that the synergetic use of data provided by these sensors allows the reliable identification of key land cover types (in primis cropped areas, bare soil areas, sparse and dense vegetated areas, forest clear cuts and burnt areas, water bodies) and their evolution over time, aimed at gathering essential information on the land cover status. In addition, it is shown that using the same repeat-pass interferometric ALOS PALSAR data pair, a Digital Elevation Model (DEM) with higher quality than the Shuttle Radar Topographic Mission (SRTM) can be generated unless the area is densely vegetated.